Functions to interpret basis sets: label-to-basis-index and basis-index-to-label #257

Activity

1. ARYAMANN7279 commented on Oct 9, 2026

   

   ARYAMANN7279

   on Oct 9, 2026

   More actions

   I'd like to take this one. Both directions come out of a single walk over shells, since conventions[(angmom, kind)] already gives the ordered labels, so there is no naming convention to invent. Working locally: a BasisFunction namedtuple plus iter_basis_functions, get_basis_function(index) and find_basis_index(label, icenter, occurrence), checked against 5 of the test fixtures where the number of walked functions matches nbasis and the first dimension of the MO coefficients. It comes to 54 lines of code but 99 with docstrings, over the 50-line guideline, so I would rather split it in two with index-to-label first. Does that split work, and should find_basis_index match a sign-flipped label like -c3 exactly, or strip the minus?
2. PaulWAyers commented on Oct 10, 2026

   

   PaulWAyers

   on Oct 10, 2026

   Member

   More actions

   • It's good to split it but 54 + 45 lines is not crazy, so it's not mandatory.
   • I'm curious what @tovrstra , @msricher and @FarnazH think, but I think that it would be nice, but inessential, to keep track of the sign convention for the basis functions.
3. PaulWAyers

   assigned

   ARYAMANN7279

   on Oct 10, 2026